VS : Systemische Physiologie - Animalische Physiologie für Bioinformatiker. Neuronenmodelle III. Modelle synaptischer Kurz- und Langzeitplastizität

Transcription

1 Bachelor Program Bioinformatics, FU Berlin VS : Systemische Physiologie - Animalische Physiologie für Bioinformatiker Synaptische Übertragung Neuronenmodelle III Modelle synaptischer Kurz- und Langzeitplastizität Martin Nawrot Neuroinformatics & Theoretical Neuroscience Institute of Biology Neurobiology

2 Outline 1. Synaptic transmission 2. Short-term synaptic plasticity 3. Long-term synaptic plasticity

3 1 Synaptic Transmission

4 1 Synaptic Transmission Excitatory PostSynaptic Current (EPSC) AP 0 Amplitude ~100 mv Duration ~1 ms stimulation of presynaptic excitatory neuron (e.g. electric or photolysis of Glu) presynaptic action potential measurement of excitatory postsynaptic current (EPSC) Experimental traces from: Boucsein et al. (2005) J Neurophysiol. 94:

5 1 Synaptic Transmission Model of PSC Simplified model for PSC: I(t) = I 0 exp (-t/τ s ) I 0 : maximum amplitude τ s : synaptic time constant from: Dyan & Abbott (2001)

6 1 Synaptic Transmission Membrane response (EPSP) to a single EPSC blackboard calculus current input to membrane I(t) = I 0 exp(-t/τ s ) impulse response of membrane A 0 exp (-t/τ m ) convolution of EPSC with membrane impulse response fcn results in postsynaptic potential PSP = PSC Synapse IR Membrane voltage output to membrane (excitatory postsynaptic potential, EPSP)

7 1 Synaptic Transmission Membrane response (EPSP) to a single EPSC synaptic weight (w)

8 2 Short-Term Synaptic Plasticity

9 2 Short term synaptic plasticity short term depression (STD) Physiology changes induced with only few presynaptic spikes Experiments: Markram H, Wang Y, and Tsodyks M (1998) PNAS. 95:5323-8

10 2 Short term synaptic plasticity short term depression (STD) Physiology Experiments: Markram H, Wang Y, and Tsodyks M (1998) PNAS. 95: Model Review: Morrison A, Diesmann M, and Gerstner W (2008) Biol Cybern. 98(6): Model changes induced with only few presynaptic spikes phenomenological model can describe this behaviour

11 Physiology changes induced with only few presynaptic spikes Experiments: Markram H, Wang Y, and Tsodyks M (1998) PNAS. 95: Short term synaptic plasticity short term potentiation (or fascilitation)

12 Physiology Model changes induced with only few presynaptic spikes phenomenological model can describe this behaviour Experiments: Markram H, Wang Y, and Tsodyks M (1998) PNAS. 95: Model Review: Morrison A, Diesmann M, and Gerstner W (2008) Biol Cybern. 98(6): Short term synaptic plasticity short term potentiation (or fascilitation)

13 2 Short term synaptic plasticity example for functional implication Olfactory system in the insect excitatory synapses between the olfactory receptor neurons (ORNs) and the antennal lobe projection neurons (PNs) exhibit short term depression (STD) A: stimulation with different odors; B: antennal nerve stimulation STD results in an adaptation of PN input current and output spike frequency with increasing input frequency. Thus, high odor concentration responses saturate at the level of PNs. Kazama & Wilson (2008) Neuron 58:

14 2 Short term synaptic plasticity example for functional implication Olfactory system in the insect excitatory synapses between the olfactory receptor neurons (ORNs) and the antennal lobe projection neurons (PNs) exhibit short term depression (STD) G: synapse model; H: Single PN responses to 2s stimulus; I: phasic-tonic response rate, trial-averaged STD can regulate response dynamics. A constant intensity odor stimulus results in a phasic-tonic PN response following odor onset. Nawrot, Krofczik, Farkhooi, Menzel (2010)

15 2 Short Term Plasticity Summary we distinguish short term potentiation at facilitating synapses and short term depression at depressive synapses short-term plasticity is induced by repeated presynaptic activation within a very short time, i.e. within tens of milliseconds (Markram et al. 1998; Thomson et al. 1993). the induced change does not persist for longer than a few hundred milliseconds (short term); original PSP amplitude is typically recovered within less than 1s.

16 3 Long-Term Synaptic Plasticity

17 3 Long term plasticity Hebbian plasticity Hebb s postulate When an axon of cell A is near enough to excite cell B or repeatedly or persistently takes part in firing it, some growth process or metabolic change takes place in one or both cells such that A s efficiency, as one of the cells firing B, is increased. Donald O. Hebb (1949)

18 3 Long term plasticity Spike Timing Dependent Plasticity (STDP) synaptic long term potentiation is induced by repetitive stimulation with positively correlated spike times of post and pre-synaptic neuron D-AP-5 (25 mm) - antagonist of NMDA subtype glutamate receptors pre post t 1 Δt = t 2 t 1 > 0 time t 2 potentiation Bi G-q, Poo M-m (1998) J Neurosci 18

19 3 Long term plasticity Spike Timing Dependent Plasticity (STDP) synaptic long term depression is induced by repetitive stimulation with negatively correlated spike times of post and pre-synaptic neuron D-AP-5 (25 mm) - antagonist of NMDA subtype glutamate receptors pre post t 2 Δt = t 2 t 1 < 0 time t 1 depression Bi G-q, Poo M-m (1998) J Neurosci 18

20 3 Long term plasticity Spike Timing Dependent Plasticity (STDP) the STDP rule describes an time-asymmetric learning rule that incorporates the phenomena of potentiation and depression depression potentiation effective time window Δt ~ 5-20ms from Dyan & Abbott (2001)

21 3 Long term plasticity STDP updating rules phenomenological models of synaptic plasticity generally operate with update rules which are central to the numeric algorithms for review see: Morrison A, Diesmann M, and Gerstner W (2008) Biol Cybern. 98(6):459-78

22 3 Long term plasticity STDP updating rules phenomenological models of synaptic plasticity generally operate with update rules which are central to the numeric algorithms Morrison A, Diesmann M, and Gerstner W (2008) Biol Cybern. 98(6):459-78

23 3 Long term plasticity Summary we distinguish two (phenomenological) types of long term synaptic changes: - Long Term Potentiation (LTP) - Long Term Depression (LTD) long-term plastic changes can be induced rapidly within 1s or less (i.e. within a rather short period, similar to the induction of short-term plasticity!) the experimentally induced change in synaptic weight are long-lasting (typically for hours or days) long term plasticity is thought to be the substance of long term memory (LTM) experimental induction protocols and underlying mechanisms are fundamentally different for short-term and long-term plastic changes We concentrated here on the well-established phenomenon of spike timing dependent plasticity (STDP) as one prominent form of establishing long-term synaptic modifications STDP is a form of Hebbian learning In STDP changes in the synaptic weight depend on the temporal relation of action potential firing in the PRE-synaptic and in the POST-synaptic neuron.

24 FIN

25 Reading Dayan & Abbott (2001) Theoretical Neuroscience: Computational and mathematical Modeling of Neural Systems, MIT Press; chapter 5.8: The Postsynaptic Conductance References Bi G-q, PooM-m (1998) Synaptic modifications in cultured hippocampal neurons: Dependence on spike timing, synaptic strength, and postsynaptic cell type. J Neurosci 18: Boucsein C, Nawrot M, Rotter S, Aertsen A, Heck D (2005) Controlling synaptic input patterns in vitro by dynamic photo stimulation. J Neurophysiol. 94: Hebb, DO (1949) The organization of behavior: a neuropsychological theory. Wiley, New York Markram H, Wang Y, and Tsodyks M (1998) Differential signaling via the same axon of neocortical pyramidal neurons. PNAS. 95: Morrison A, Diesmann M, Gerstner W. (2008) Phenomenological models of synaptic plasticity based on spike timing. Biol Cybern. 98(6): Nawrot MP, Krofczik S, Farkhooi F, Menzel R (2010) Temporal Dynamics of odor processing in the insect antennal lobe. Frontiers in Neuroscience (submitted) Thomson AM, Deuchars J, West DC (1993) Large, deep layer pyramid-pyramid single axon EPSPs in slices of rat motor cortex display paired pulse and frequency-dependent depression, mediated presynaptically and self-facilitation, mediated postsynaptically. J Neurophysiol Dec;70(6):

26 4 Unsupervised Supervised - Reward-Based Learning for all Hebbian learning rules only variables that are locally available at the synapse can be used to change the synaptic weight (unsupervised) teaching signal provides non-local (global) control reward-based (reinforcement) learning allows for self-controlled systems - not covered in WS 09/10